skills.rest
TrayMachi
TrayMachiCOMMUNITY
Explore all skills by this author →
WebsiteGitHub

caveman

Compresses model responses into terse caveman-style prose while preserving technical accuracy.

Updated Mar 30, 2026
Search Tags:#token efficiency#output formatting#token compression#classical chinese#caveman mode#concise responses#prompt style
One-click install
npx skills add https://github.com/TrayMachi/dotfiles --skill caveman-traymachi
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: caveman
Source: https://github.com/TrayMachi/dotfiles/tree/main/agents/skills/caveman
Command: npx skills add https://github.com/TrayMachi/dotfiles --skill caveman-traymachi

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Long, filler-heavy AI responses waste tokens and slow down reading. This Skill cuts output token usage by roughly 65-75% by stripping articles, pleasantries, and hedging while keeping every technical detail, code block, and error string exact. ## Core Features & Use Cases - Six intensity levels: lite, full (default), ultra, plus three Classical Chinese (wenyan) variants for up to 80-90% character reduction. - Session persistence: the mode stays active across all turns until explicitly stopped with "stop caveman" or "normal mode". - Auto-clarity fallback: automatically reverts to normal prose for security warnings, irreversible-action confirmations, and ambiguous multi-step sequences, then resumes compression. - Use Case: During a long debugging session, enable ultra mode so every explanation arrives as compact fragments like "Inline obj prop → new ref → re-render. useMemo.", cutting reading time without losing precision. ## Quick Start Ask the assistant to talk like caveman or invoke /caveman to enable compressed responses for the rest of the session.

Frequently Asked Questions about caveman

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make AI responses shorter and use fewer tokens?▼

Enable caveman mode by saying "talk like caveman" or invoking /caveman. It drops articles, filler, and pleasantries while keeping code, error strings, and technical terms exact, cutting output tokens by roughly 65-75%.

How do I change the caveman compression level?▼

Invoke /caveman with a level argument: lite, full, ultra, wenyan-lite, wenyan-full, or wenyan-ultra. Full is the default; ultra adds abbreviations and arrows, while wenyan levels compress into Classical Chinese.

Does caveman mode stay active for the whole conversation?▼

Yes, the mode persists across every response in the session once enabled. It only turns off when you say "stop caveman" or "normal mode", and the chosen level persists until changed.

When does caveman mode switch back to normal prose?▼

It automatically drops to normal prose for security warnings, irreversible-action confirmations, ambiguous multi-step sequences, and when you repeat a question or ask for clarification. Compression resumes after the clear part.

Does caveman mode change code blocks or error messages?▼

No. Code blocks, function names, API names, and error strings are never abbreviated or altered. Only surrounding prose is compressed, and code, commits, and PR descriptions are written normally.

Related Skills

Explore More Software Engineering Skills→
MoofonLiMoofonLi
2

caveman

Compresses AI responses into terse caveman-style prose while preserving technical accuracy.

Community
Intermediate
uplbtoolsuplbtools
66

room-tba-agent-workflow

Guide coding workflow for the Room TBA project with scoped edits and verification.

Official
Intermediate
gonzalezpazmonicagonzalezpazmonica
47

headroom-optimization

Compress AI context windows using a 5-phase framework to reduce token usage.

Community
Intermediate
intrafindintrafind
22

add-app

Scaffold iHub AI application configuration files in JSON with Zod schema validation.

Official
Basic